From Pages to Practice

By Bhavna Seth, M.D.

Published December 12, 2018


Ms. Jones is a 65-year-old woman in the intensive care unit (ICU) who has been intubated for the last 72 hours due to respiratory failure and septic shock from pneumonia. During rounds, your team runs through the daily ICU patient checklist. For the question about whether Ms. Jones meets indications for proton pump inhibitor (PPI) prophylaxis, you discuss the possible benefits and adverse effects of these agents, including an increased risk of pneumonia and Clostridium difficile infection. Is the evidence clear that PPI prophylaxis will improve Ms. Jones’ outcome?

To examine this question, investigators conducted a European, multicenter, parallel-group, blinded trial (the SUP-ICU trial) in 3298 ICU patients with at least one risk factor for clinically important gastrointestinal (GI) bleeding including shock, use of anticoagulant agents, renal-replacement therapy, mechanical ventilation (expected to last >24 hours), history of liver disease, or coagulopathy. The primary outcome was death by 90 days after randomization. The secondary outcome was a composite of clinically important events in the ICU, including GI bleeding, new-onset pneumonia, C. difficile infection, or acute myocardial ischemia. GI bleeding was defined as overt GI bleeding and at least one of the following four features within 24 hours of the event (in the absence of other causes): a spontaneous decrease in systolic blood pressure, mean arterial pressure, or diastolic blood pressure of 20 mm Hg or more; initiation of treatment with a vasopressor or a 20% increase in vasopressor dose; a decrease in hemoglobin of at least 2 g per deciliter; or transfusion of two or more units of packed red cells.

Mortality at 90 days did not differ significantly between the pantoprazole and placebo groups (31.1% vs. 30.4%; relative risk, 1.02; 95% CI, 0.91 to 1.13; P = 0.76). Similarly, the number of patients who developed one or more clinically important events in the ICU did not differ significantly (21.9% vs. 22.6%; RR, 0.96; 95% CI, 0.83 to 1.11). However, 41 patients in the pantoprazole group had clinically important GI bleeding, as compared with 69 patients in the placebo group (2.5% vs. 4.2%; no P value because of the lack of adjustment for multiple comparisons.)

These results are similar to those of a recent meta-analysis, in which no significant differences were found in rates of death or infectious complications between PPI prophylaxis and placebo. Strengths of the current trial were its large sample size and pragmatic design. However, the lack of reported serious adverse events may reflect underreporting, as had likely occurred in the ADRENAL trial for glucocorticoids in septic shock. Further, it was not possible to distinguish between stress ulcers and other causes of GI bleeding because diagnostic endoscopies were not mandated. 

NEJM Deputy Editor Dr. Allan H. Ropper wonders if the results of this study will markedly reduce use of GI prophylaxis because of the difference between groups in clinically important GI bleeding, but adds that it should at least make us cautious in their use. Editorialists Dr. Barkun, from McGill University in Montreal and Dr. Bardou from Dijon, France conclude that given the likely low incidence of clinically important upper GI bleeding in the ICU, and only a 1.7-percentage-point difference in bleeding risk between groups, PPI prophylaxis should be “reserved for seriously ill patients who are at high risk for this complication.” They note, however, that the definition of high risk of bleeding has not been standardized.

For Ms. Jones, your team decided to add pantoprazole prophylaxis, given her acute septic shock state and increased risk of stress ulcer-related bleeding.

Browse more From Pages to Practice »

Bhavna Seth, Resident at Boston Medical Center
Bhavna Seth, Resident at Boston Medical Center
NEJM Knowledge+